Object Recognition Device Pose-Based Possession Estimation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing object detection techniques struggle to accurately recognize a person's possession in images, especially when the possession is not clearly visible due to occlusion or low resolution.

Innovation Solution

An object recognition device that estimates a person's possession based on their pose in an image, combines this estimation with object detection results, and adjusts weights to improve recognition accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If normal object detection technique is used, then detection process is simple, but detection accuracy of possession is reduced when possession is not clearly visible

Engineering Contradiction:

Improvedetection accuracy of possessionVSAvoiddetection process complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent combines multiple detection approaches: pose-based estimation (using human body pose information to infer possession location) and image-based object detection (using visual recognition algorithms). This merging allows the system to maintain high detection accuracy even when the possession is not clearly visible in the image, by cross-validating results from both methods.

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

The patent introduces pose information as an intermediary element that bridges the gap between the person and the possession. By detecting the person's pose and using it to estimate possession location, the system can infer possession information even when the possession itself is not directly visible in the image.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Reliability

If pose-based estimation is used, then recognition accuracy improves when possession is hidden, but computational complexity increases

Engineering Contradiction:

Improverecognition accuracy of hidden possessionVSAvoidcomputational process complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent applies partial action by using pose-based estimation only when the possession is not clearly detected through image analysis. The system first attempts image-based detection, and only resorts to pose-based estimation when necessary, thereby reducing overall computational complexity while maintaining high reliability for difficult cases.

Inventive Principle:

Principle #16Partial or excessive action

Solution Approach 2:

The patent changes the detection parameters dynamically based on image conditions. When the possession is clearly visible, the system uses standard object detection parameters. When the possession is hidden or occluded, the system switches to pose-based estimation parameters, optimizing computational efficiency for different scenarios.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS12307707B2Object recognition device, object recognition method, and recording medium

Publication Date: 2025.05.20 NEC CORP

AI summary

In an object recognition device, an estimation means estimates a possession of the person based on a pose of a person in an image. An object detection means detects an object from surroundings of the person in the image. A weighting means sets weights with respect to an estimation result of the possession and a detection result of the object based on the image. A combination unit combines the estimation result of the possession and the detection result of the object by using the weights being set, and recognizes the possession of the person.